Multiple client issues that went undetected by On Call team (but reported by CEE)- tracked via QUAY-2033.

It was discovered that the unhealthy host count for clair workers alarm was firing in cloudwatch however no pagerduty alert was configured for it (discovered by Evan).  Need to fix this so that On Call is notified when Clair worker count is unhealthy.

Prometheus had a similar issue recently.